Facial Recognition with Simulated Occlusions
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Facial recognition systems face challenges in accurately identifying faces from images captured by different cameras, especially when faces are partially covered by masks, sunglasses, or other obstructions, leading to reduced recognition accuracy.
Innovation Solution
A computing device uses a trained model to associate images of partially covered faces with a reference set of unmasked faces, expanded through simulated face coverings, ensuring accurate matching by comparing masked and unmasked faceprints, and selecting high-quality images for reference sets to improve detection accuracy.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If facial recognition is performed on images from different cameras, then the system can process diverse image sources, but recognition accuracy deteriorates due to quality variations
Solution Approach 1:
The system performs preliminary actions by generating simulated face coverings and expanding the training dataset before actual recognition. This pre-processing allows the model to learn from diverse conditions (different cameras, lighting, occlusions) and improve its ability to handle quality variations in real recognition tasks
Solution Approach 2:
The system changes parameters by applying simulated face coverings (masks, sunglasses, hats) to training images, thereby modifying the input data characteristics. This allows the model to learn robust features that remain consistent across different camera qualities and occlusion conditions
2Adaptability or versatility
If face coverings are present in images, then real-world recognition scenarios are better represented, but recognition accuracy deteriorates due to obscured facial features
Solution Approach 1:
The system creates copies of original face images with simulated coverings applied. These synthetic covered-face images serve as training examples, allowing the model to learn recognition patterns for covered faces without requiring actual covered-face photographs, thereby maintaining accuracy while improving real-world adaptability
Solution Approach 2:
The system performs preliminary action by pre-generating images with simulated face coverings during the training phase. This allows the recognition model to learn from diverse occlusion scenarios before encountering real covered faces during deployment, improving its ability to handle real-world scenarios with face coverings
Data Source
AI summary
Techniques are disclosed for determining the presence of a particular person based on facial characteristics. For example, a device may include a first image in a reference set of images based on determining that a face shown in the first image is not covered by a face covering. A trained model of the device may determine a first set of characteristics from the first image, whereby the trained model is trained utilizing simulated face coverings to match a partially covered face of a particular person with a non-covered face of the particular person. The device may also determine a second set of characteristics associated with a face of a second person based on a second image. The trained model may then determine a score corresponding to a level of similarity between both sets of characteristics, and then determine whether the first person is the second person based on the score.